summary - Uses phylodynamic rooting plus epidemic simulation to argue that the two earliest-observed SARS-CoV-2 genomic lineages (A and B) reflect at least two independent zoonotic introductions into humans, rather than in-human diversification from one index case: lineage B likely jumped to humans around mid-November 2019, with lineage A following within weeks. Simulations reproduced the observed phylogenetic topology (two large basal polytomies) in only ~0.5% of single-introduction scenarios; a molecular-clock test rejected a single-ancestor lineage-A haplotype (Bayes factor 48.1) in favor of a second, separate introduction.
relevance_note - The primary quantitative case that the pandemic began with 2+ independent spillovers rather than one - harder to square with a single lab-accident origin story (which would more naturally predict one index lineage), easier to square with repeated wildlife-trade exposure at a market. Genomic/sequence-divergence analysis over the shared SARS-CoV-2 GISAID sequence collection (recurring-dataset note), but a distinct primary artifact/method (lineage-defining mutations + molecular clock) from the furin-site papers in this slice and from slice 1’s market-geolocation paper.